A Message from the Editor

Driving Signature Analysis for Auto-Theft Recovery

Abstract: Autotheft is a crime that can be mitigated using artificial intelligence as a scientific approach. In this case, we assess the drivers driving pattern using both deep neural network and swarm intelligence algorithms. From the analysis we are able to obtain the driving signature of the driver which can be associated with the vehicle. The vehicle is then tracked and monitored. Next, a deviation from the usual driving signature of the owner or assigned driver would signify a possible instance of autotheft. Subsequently, the vehicle can be traced and reclaimed by the owner. The algorithms are evaluated based on their performance in analysing the datasets bearing variable features. The variations in features enable us to verify the efficacy and accuracy levels of the various algorithms that are used in the study. The metrics used for evaluation are the Mean Squared Error and the F1 Score for precision, accuracy and recall functionality.

Keywords: Deep learning, swarm intelligence, driving signature, intelligent transportation system.

Received April 5, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/1

Full text

An Efficient Parallel Version of Dynamic Multi-Objective Evolutionary Algorithm

Abstract: Multi-Objective Optimization Evolutionary Algorithms (MOEAs) belong to heuristic methods proposed for solving Multi-objective Optimization Problems (MOPs). In fact, MOEAs search for a uniformly distributed, near-optimal, and near-complete Pareto front for a given MOP. However, several MOEAs fail to achieve their aim completely due to their fixed population size. To overcome this shortcoming, Dynamic Multi-Objective Evolutionary Algorithm (DMOEA) [20] was proposed. Although DMOEA has the distinction of dynamic population size, it still suffers from a long execution time. To deal with the last disadvantage, we have proposed previously a Parallel Dynamic Multi-Objective Evolutionary Algorithm (PDMOEA) [10] to obtain efficient results in less execution time than the sequential counterparts, in order to tackle more complex problems. This paper is an extended version of [10] and it aims to demonstrate the efficiency of PDMOEA through more experimentations and comparisons. We firstly compare DMOEA with other multi-objective evolutionary algorithms Non-Dominated Sorting Genetic Algorithm (NSGA-II) and Strength Pareto Evolutionary Algorithm (SPEA-II), then we present an exhaustive comparison of PDMOEA versus DMOEA and discuss how the number of used processors influences the efficiency of PDMOEA. As experimental results, PDMOEA enhances DMOEA in terms of three criteria: improving the objective space, minimizing the computational time, and converging to the desired population size. Finally, the paper establishes a new formula relating the suitable number of processes, required in PDMOEA, and the number of necessary generations to converge to the optimal solutions.

Keywords: Multi-objective problems, pareto front, multi-objective evolutionary algorithms, dynamic MOEA, parallel DMOEA.

Received April 10, 2022; accepted April 28, 2022
https://doi.org/10.34028/iajit/19/3A/2

Full text

A Schema-Free Instance Matching Algorithm Based on Virtual Document Similarity

Abstract: With the continuous development of semantic web, especially of the web of data, several knowledge bases expressed by ontologies are independently created and added to the Linked Open Data (LOD) cloud, on a daily basis. A major challenge for the LOD paradigm is to discover resources that refer to the same real-world object, in order to interlink web resources and hold large scale data integration and sharing. In this context, instance matching is a promising solution. It aims to link co-referent instances belonging to heterogeneous knowledge bases with owl: same as links. Several state-of-the-art existing approaches addressing this issue are based on the prior schema-level matching's, which does not avoid the limitation of heterogeneity at the property-level. In this paper, we propose a schema-free, scalable and efficient instance matching approach that is independent from matching results at the schema-level. We transform the instance matching problem to a document similarity problem and we solve it by a Clustering technique that uses an Ascendant Hierarchical Clustering algorithm to group similar instances in the same clusters. Furthermore, we design multiple validating patterns that use some structural information to validate obtained mappings and eliminate wrong ones. Experiments on instance matching track from Ontology Alignment Evaluation Initiative (OAEI) show that our approach gets prominent results compared to several participating systems in OAEI’2019, OAEI’2020 and OAEI’2021.

Keywords: Ontology, LOD, instance matching, ascendant hierarchical clustering, OAEI.

Received April 9, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/3

Full text

Assesing the Stability and Selection Performance of Feature Selection Methods Under Different Data Complexity

Abstract: Our study aims to investigate the stability and the selection accuracy of feature selection performance under different data complexity. The motivation behind this investigation is that there are significant contributions in the research community from examining the effect of complex data characteristics such as overlapping classes or non-linearity of the decision boundaries on the classification algorithm's performance; however, relatively few studies have investigated the stability and the selection accuracy of feature selection methods with such data characteristics. Also, this study is interested in investigating the interactive effects of the classes overlapped with other data challenges such as small sample size, high dimensionality associated with irrelevant features, and imbalance classes to provide meaningful insights into the root causes for feature selection methods misdiagnosing the relevant features among different real-world data challenges. This analysis will be extended to real-world data to guide the practitioners and researchers in choosing the correct feature selection methods that are more appropriate for a particular dataset. Our study outcomes indicate that using feature selection techniques with datasets of different characteristics may generate different subsets of features under variations to the training data showing that small sample size and overlapping classes have the highest impact on the stability and selection accuracy of feature selection performance, among other data challenges that have been investigated in this study. Also, in this study, we will provide a survey on the current state of research in the feature selection stability context to highlight the area that requires more attention for other researchers.

Keywords: Stability of feature selection, class overlapping, data characteristics, complex data.

Received April 10, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/4

Full text

Automotive Embedded Systems-Model Based Approach Review

Abstract: The evolution of transforming from an electrical mechanical engineering discipline to a combination of software and electrical/mechanical engineering establishes software as a crucial technology. The current complex automotive system is the product of growth of embedded software. As a result, automotive industry focuses on a new trend Model based development rather than traditional method where software is handwritten in Assembly code or C language. This paper presents a review of the use of Model based Development to accelerate development process of embedded control systems and technologies. The paper also presents a review of the tools used to support Model-Based Development (MBD) from functional requirements to automated testing and Model based testing process

Keywords: Model-based development, automotive embedded systems, embedded software, automotive industry.

Received March 24, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/5

Full text

Densely Convolutional Networks for Breast Cancer Classification with Multi-Modal Image Fusion

Abstract: Breast cancer is the main health burden worldwide. Cancer is located in the breast, starts when the cell grows under control and begins as in-situ carcinoma and when spread into other parts known as invasive carcinoma. Breast cancer mass can early be found by image modality when discovering mass early can easily diagnose and treated. Multimodalities used for the classification of breast cancer Such as mammography, ultrasound, and Magnetic resonance imaging. Two types of fusion are used earlier fusion and later fusion. Early fusion it’s a simple relation between modalities while later fusion gives more interest to fusion strategy to learn the complex relationship between various modalities as a result, can get highly accurate results when using the later fusion. When combining two image modalities (mammography, ultrasound) and using an excel sheet containing the age, view, side, and status attribute associated with each mammographic image using DenseNet 201 with Layer level fusion strategy as later fusion by making connections between the various paths and same path by using Concatenated layer. Fusing at the feature level achieves the best performance in terms of several evaluation metrics (accuracy, recall, precision area under the curve, and F1 score) and performance.

Keywords: Breast cancer/classification, deep learning, densenet, diagnostic imaging, multimodal imaging.

Received April 3, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/6

Full text

DWT and LBP Map Based Feature Descriptors for Face Recognition in Harsh Light Variations

Shekhar Karanwal

Graphic Era University (Deemed),

India

This email address is being protected from spambots. You need JavaScript enabled to view it. 

Abstract: In [11] Karanwal et al. provided enhancements to three descriptors for Face Recognition under illumination variations. The three descriptors for which enhancements are done are Local Binary Pattern (LBP), Horizontal Elliptical LBP (HELBP) and Median Binary Pattern (MBP). By deploying Two Dimensional DWT (2D-DWT) (utilizing haar at level 1) before features extraction of LBP, HELBP and MBP, the enhancements are made. These improved ones outperforms the original descriptors comprehensively. After careful analyzing the work proposed in [11] it has been observed that even after image pre-processing, histograms of LBP, HELBP and MBP unable to capture the efficient information to declare as the robust descriptors in light variations. In the proposed work it has been observed and implemented that map feature of LBP, HELBP and MBP (after image pre-processing by 2D-DWT) yields much better accuracy than the histogram based descriptors. The three proposed descriptors are 2D-DWT+LBP_map, 2D-DWT+HELBP_map, 2D-DWT+MBP_map. These map features full and completely outperform its respective histogram features & these are LBP_hist, 2D-DWT+LBP_hist, HELBP_hist, 2D-DWT+HELBP_hist, MBP_hist and 2D-DWT+MBP_hist. Among all it is 2D-DWT+HELBP_map feature which yields best results. The feature compression is fulfilled by the usage of Fishers Linear Discriminant Analysis (FLDA) and classification was done from Support Vector Machines (SVMs). For experiments Yale B (YB) and Extended Yale B (EYB) datasets are used.

Keywords: LBP, HELBP, MBP, (2D-DWT), histogram feature, map feature.

Received March 20, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/7

Full text

Stacknet Based Decision Fusion Classifier for Network Intrusion Detection

Abstract: Network intrusion is a subject of great concern to a variety of stakeholders. Decision fusion (ensemble) models that combine several base learners have been widely used to enhance detection rate of unauthorised network intrusion. However, the design of such an optimal decision fusion classifier is a challenging and open problem. The Matthews Correlation Coefficient (MCC) is an effective measure for detecting associations between variables in many fields; however, very few studies have applied it in selecting weak learners to the best of the authors’ knowledge. In this paper, we propose a decision fusion model with correlation-based MCC weak learner selection technique to augment the classification performance of the decision fusion model under a StackNet strategy. Specifically, the proposed model sought to improve the association between the prediction accuracy and diversity of base classifiers. We compare our proposed model with five other ensemble models, a deep neural model and two stand-alone state-of-the-art classifiers commonly used in network intrusion detection based on accuracy, the Area Under Curve (AUC), recall, precision, F1-score and Kappa evaluation metrics. The experimental results using benchmark dataset KDDcup99 from Kaggle shows that the proposed model has a identified unauthorised network traffic at 99.8% accuracy, Extreme Gradient Boosting (Xgboost) (97.61%), Catboost (97.49%), Light Gradient Boosting Machine (LightGBM) (98.3%), Multilayer Perceptron (MLP) (97.7%), Random Forest (RF) (97.97%), Extra Trees Classifier (ET) (95.82%), Different decision (DT) (96.95%) and, K-Nearest Neighbor (KNN) (95.56), indicating that it is a more efficient and better intrusion detection system.

Keywords: Network intrusion detection, stacknet, ensemble learning classifier.

Received April 8, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/8

Full text

Retention Contracts under Partial Information Electoral Competition Case Study

Abstract: This study copes with a class of principal-agent problems where information asymmetry represents an important characteristic. The paper examines the relationship between the principal and agents. The principal has to perform two agents’ screening and discipline tasks. To complete his duties, the principal lacks complete information concerning the agents’ behavior and rarely has partial information regarding the failure or success of launched tactics, alliances, rationalization, etc. We analyze the type of retention contracts (implicit) used by the principal to replace or retain agents. Consistent with literature findings, we demonstrated that agents could be extremely active in showing their competencies; the relationship between dismissal and bad performance is invalid; and occasionally, the principal dismisses qualified agents. Then we determined the rules under which electorates urge political parties to acquire information and choose optimal policies from the voter’s viewpoint.

Keywords: Retention contracts, moral hazard, principal-agent problem, electoral competition.

Received March 30, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/9

Full text

The Intrusion Detection System by Deep Learning Methods: Issues and Challenges

Abstract: Intrusion Detection Systems (IDS) are one of the major research application problems in the computer security domain. With the increasing number of advanced network attacks, the improvement of the traditional IDS techniques become a challenge. Efficient ways and methods of identifying, protecting, and analyzing data are needed. In this paper, a comprehensive survey on the application of Machine Learning (ML) and Deep Learning (DL) methods on the IDS to increase detection accuracy and reduce error rate is proposed. The recent research papers that have been published between 2018 and 2021 in the area of applying ML and DL in the IDS are analyzed and summarized. Four main analyzing aspects are presented as follows: (1) IDS concepts and taxonomy. (2) The strength and weaknesses of ML and DL methods. (3) IDS benchmark datasets. (4) Comprehensive review of the most recent articles that used ML and DL to improve IDS with highlighting the strengths and weaknesses of each work. Based on the analysis of the literature review papers, a framework for the application of ML and DL in the IDS is proposed. Finally, the current limitations are discussed and future research directions are provided.

Keywords: Artificial intelligence, dataset, deep learning, intrusion detection, machine learning, security.

Received April 2, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/10

Full text

Analyzing the Effect of Driving Speed on the Performance of Roundabouts

Abstract: Roundabouts are introduced at intersections to improve traffic flow and enhance safety. Nevertheless, studies and field observations demonstrated that roundabouts, due to their special design, could significantly affect the efficiency of the overall traffic network, especially in the case of increased traffic volumes. For this reason, researchers and practitioners have conducted several studies to alleviate the negative impact of saturated traffic. In these studies, different characteristics of traffic flow and roundabout topologies are analyzed to show their impact on the overall performance. In this paper, we present two simulation studies to investigate the effect of driving speed on the performance of roundabouts with different geometrical characteristics. The results from the two case studies indicate that speed control and the distribution of traffic volumes on the arms of the roundabout are two important factors that affect the performance of roundabouts. Moreover, the results also show that the driving speed factor correlates with roundabouts' geometrical characteristics. Further, the individual driving behavior plays a major role in the performance of roundabouts.

Keywords: Roundabout, traffic simulation, congestion.

Received April 9, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/11

Full text

The Critical Feature Selection Approach using Ensemble Meta-Based Models to Predict Academic Performances

Abstract: In this work, machine learning techniques are deemed to predict student academic performances in their historical performance of Final Grades (FGs). Acceptance of Technology enabled the teaching-learning processes, as it has become a vital element to perceive the goal of academic quality. Research is improving and growing fast in Educational Data Mining (EDM) due to many students' information. Researchers urge to invent valuable patterns about students' learning behavior using their data that needs to be adequately processed to transform it into helpful information. This paper proposes a prediction model of students' academic performances with new data features, including student's behavioral features, Psychometric, family support, learning logs via e-learning management systems, and demographic information. In this paper, data collection and pre-processing are firstly conducted following the grouping of students with similar patterns of academic scores. Later, we selected the applicable supervised learning algorithms, and then the experimental work was implemented. The performance of the student's predictive model assessment is comprised of three steps: First, the critical Feature selection approach is evaluated. Second, a set of renowned classifiers are trained and tested. Third, ensemble meta-based models are improvised to boost the accuracy of the classifier. Subsequently, the present study is associated with the solutions that help the students evaluate and improve their academic performance with a glimpse of their historical grades. Ultimately, the results were produced and evaluated. The results showed the effectiveness of our proposed framework in predicting students' academic performance.

Keywords: Educational data mining, students' prediction, machine learning, ensemble meta-based models, feature selection.

Received April 11, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/12

Full text

Echo State Network Optimization using Hybrid-Structure Based Gravitational Search Algorithm with Square Quadratic Programming for Time Series Prediction

Abstract: The Echo-State Network (ESN) is a robust recurrent neural network and a generalized form of classical neural networks in time-series model designs. ESN inherits a simple approach for training and demonstrates the high computational capability to solve non-linear problems. However, input weights and the reservoir's internal weights are pre-defined when optimizing with only the output weight matrix. This paper proposes a Hybrid Gravitational Search Algorithm (HGSA) to compute ESN output weights. In Gravitational Search Algorithm (GSA), Square Quadratic Programming (SQP) is united as a local search strategy to raise the standard GSA algorithm's efficiency. Later, an HGSA-SQP and the validation data set to establish the relation configuration of the ESN output weights. Experimental results indicate that the proposed configuration of HGSA-SQP-ESN is more efficient than the other conventional models of ESN with the minimum generalization error.

Keywords: Echo state network, hybrid gravitational search algorithm, network configuration optimization, time series prediction.

Received April 10, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/13

Full text

Random Walk Generation and Classification Within an Online Learning Platform

Abstract: Advancements in technology have introduced new approaches in teaching and learning processes. Machine learning algorithms analyse and recognize patterns of data and subsequently become able to make reasonable decisions. In playing complex games, such as chess and go, machine learning algorithms have even already outperformed humans. This paper presents a software platform ‘Discimus_RW’ that introduces a novel approach for teaching, learning, and researching random walk theory and getting hands-on experience in machine learning. Random walk theory represents the foundations of many fundamental processes, including the diffusion of substances in solvents, epidemics’ spread, and financial markets’ development. ‘Discimus_RW’ is composed of three main features: 1. Random walk generation using mathematical Equations, 2. Random walk classification using supervised learning algorithms, and 3. Random walk visualization. A few users who explored ‘Discimus_RW’ showed an interest and positive feedback that assured the experiential learning experience achieved using this software, which will therefore reinforce random walk teaching and learning.

Keywords: Random walk, machine learning, experiential learning.

Received April 8, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/14

Full text

The Jordanian Universities Experience in Integrating Online Learning and its Quality Assurance

Abstract: University educational institutions in Jordan, in general, have quickly dealt with emergency situations facing education, such as the Corona pandemic, in which they needed to move from traditional learning to online learning (in its fully and blended forms) during the previous period. This led the public and private educational institutions in Jordan to implement the executive action plan that was prepared by the Ministry of Higher Education and Scientific Research (MoHESR) for integrating online learning into the higher education system and reflect it on their programs in order to keep pace with developments at the local, regional and international levels. This study includes a specification of the components relevant to online learning management system, and aims to determine the percentages of achievement at the level of fully online learning and blended learning for all components of the online learning management system in public and private Jordanian universities and also determine the percentages of achievement for each one of these components at the university level, and the most important actions that should be taken by universities to achieve them. The study also aims to benefit the integration of online learning within the Jordanian higher education system in an effective manner that achieves high levels of educational quality for online learning in its both forms and ensure the desired shift in the performance of Jordanian higher education institutions and the quality of its output, keeping pace with global developments in this field. The study concludes with a set of results that will benefit decision-makers in the Accreditation and Quality Assurance Commission for Higher Education Institutions (AQACHEI), MoHESR, and Jordanian higher education institutions.

Keywords: Online Learning, Jordanian Universities, Fully Online Learning, Blended Learning.

Received April 6, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/15

Full text

The Role of Artificial Intelligence Abilities in Library Services

Abstract: Artificial intelligence applications are one of the most important modern technologies that have emerged in recent times, which have shaped a major change in the functions of institutions, especially the scope of libraries and information. This study was described the role of Artificial Intelligence Applications of knowledge management in different information organizations, as well as to identify the reality of employing artificial intelligence applications in libraries to enrich the Knowledge Management, exposing the relationship between artificial intelligence applications and its ability to develop technical and administrative processes in libraries of Knowledge Management and Knowing the challenges facing the libraries to apply the artificial intelligence applications.

The research is based on descriptive method by content analysis of literature review by listing the most important and related published Arabic and foreign literature it was token the studies that are focused in Artificial Intelligence Applications of knowledge management in information organizations.

The study conclude that the most important results of which are: There are a lot of libraries that apply artificial intelligence technology in their services, whether technical in the provision of services, references and others, which undoubtedly facilitated the ease of retrieval and search for the users, the available studies discussed various aspects of the reality that this new application plays in facilitating library services, and most studies agree on that to facilitate services and their quality intelligence and the application of artificial intelligence requires the availability of a set of basic ingredients and requirements, such as strong technical equipment, and qualified human resources capable of using and developing this technology. The study recommended the necessity for libraries and information centers to strive to keep pace with the changes of artificial intelligence and to make the right investment for knowledge management, increasing field studies to explore the requirements of artificial intelligence technologies and diversity in the services provided by artificial intelligence in the library and activating them effectively.

 Keywords: Artificial intelligence, information organizations, knowledge management.

Received April 6, 2022; accepted April 28, 2022

https://doi.org/10.34028/iajit/19/3A/16

Full text